Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Locally reverse combined sparse representation object tracking method of multi-template space-time association

A spatiotemporal correlation and joint sparse technology, applied in computer parts, character and pattern recognition, instruments, etc., can solve problems such as reducing computational complexity

Active Publication Date: 2017-06-20
UNIV OF ELECTRONIC SCI & TECH OF CHINA
View PDF8 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] Based on the above technical problems, the present invention provides a local anti-joint sparse representation target tracking method based on multi-template spatio-temporal correlation, which aims to mine the spatio-temporal correlation information between different templates, so as to solve complex problems such as local occlusion, illumination changes, and scale changes. The tracking problem in the scene, while reducing the computational complexity

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Locally reverse combined sparse representation object tracking method of multi-template space-time association
  • Locally reverse combined sparse representation object tracking method of multi-template space-time association
  • Locally reverse combined sparse representation object tracking method of multi-template space-time association

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0108] attached Figure 5 is the experimental result on the "Football" sequence. In this video, the target rotates during motion, is partially blocked by surrounding objects, and the background is cluttered. Using the local anti-joint sparse representation target tracking method of multi-template spatio-temporal correlation adopted in the present invention, the similar parts of each target template can be more fully mined, so that candidate targets can be screened by using the unoccluded target image block representation model. Experiments show that the method is robust to local variations of objects.

Embodiment 2

[0110] attached Figure 6 It is the experimental result on the "Quadrocopter" sequence. The target in this video is moving at a relatively fast speed, and the resolution of the target is low due to infrared imaging. In the tracking process, the target motion model is used to predict the possible position and state of the target in the next frame, the candidate target selection problem is modeled as the contribution size of the reconstructed target template, and multiple candidate targets are scored, and the final decision is made jointly The experimental results show that this method can well solve the problem of fast movement of the target and the problem of low resolution.

Embodiment 3

[0112] attached Figure 7 It is the experimental result on the "Trellis" sequence. The object appearance model is affected by illumination changes, scale changes, and object rotation. The measurement method based on cosine distance has the characteristics of strong robustness to illumination changes, and the target in this video is gradually changing, so it can make full use of the nearest neighbor template to select candidate targets, and multi-task learning on multiple templates can prevent The introduction of wrong templates affects the cumulative error caused by tracking. Experimental results prove that the present invention can overcome the illumination change, size change and rotation of the target.

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention relates to a locally reverse combined sparse representation object tracking method of multi-template space-time association, belongs to the field of machine vision and mode identification, and aims at digging out space-time association information among templates, solving the problem of object tracking in a complex scene, and reducing the computing complexity. Under a particle filtering framework, a candidate object is collected and a local image block is divided to construct an over-complete dictionary; an object template sample set is initialized and updated, and the templates are labeled according to obtained frame numbers; a sequential weight matrix is established to guide sequential prior information of sparse representation; the dictionary is used to carry out weighted combined sparse representation on all samples in the object template set, a combined sparse constraint in an object function is replaced with a parameter included Gaussian function to optimize problem solution, and a sparse representation coefficient matrix is obtained; and according to row continuity and weight of distribution of non-zero elements in the coefficient matrix, a significance score of the candidate object is obtained, and stable object tracking is realized.

Description

technical field [0001] The invention relates to the fields of machine vision and pattern recognition, in particular to a multi-template spatio-temporal correlation local anti-union sparse representation target tracking method. Background technique [0002] Object tracking is an important research direction in the field of computer vision, and it has been widely used in public transportation intelligent monitoring systems, security systems, human-computer interaction and other fields. Due to the complexity of the tracking scene, the target appearance model often undergoes large and unpredictable changes due to the interference of external objects and its own deformation, which makes it a huge challenge to achieve continuous, stable and robust target tracking. In order to solve the above problems, it is necessary to establish an appropriate representation model for the target to be tracked to overcome complex situations such as illumination changes, scale changes, and local oc...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G06K9/00G06K9/62
CPCG06V20/48G06F18/22
Inventor 彭真明李美惠陈科潘翯陈颖频王晓阳孙伟嘉任丛雅旭卓励然蒲恬张萍
Owner UNIV OF ELECTRONIC SCI & TECH OF CHINA
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com